Free
Education  |   January 2003
Multicenter Randomized Comparison of the Efficacy and Safety of Xenon and Isoflurane in Patients Undergoing Elective Surgery
Author Affiliations & Notes
  • Rolf Rossaint, M.D.
    *
  • Matthias Reyle-Hahn, M.D.
  • Jochen Schulte am Esch, M.D.
  • Jens Scholz, M.D.
    §
  • Philippe Scherpereel, M.D.
  • Benoit Vallet, M.D.
    #
  • Francesco Giunta, M.D.
    **
  • Monica Del Turco, M.D.
    ††
  • Wilhelm Erdmann, M.D.
    ‡‡
  • Alfons F. Hammerle, M.D.
    ∥∥
  • Peter Nagele, M.D.
    ***
  • * Professor and Chairman, † Staff Anesthesiologist, Department of Anesthesiology of the University Hospital, RWTH Aachen, Germany. ‡ Professor and Chairman, § Professor, Department of Anesthesiology of the University Hospital Eppendorf, Hamburg, Germany. ∥ Professor and Chairman, # Professor, Department of Anesthesiology and Intensive Care, University Hospital of Lille, Lille, France. ** Professor and Chairman, †† Staff Anesthesiologist, Department of Anesthesiology and Intensive Care, University of S. Chiara Hospital, Pisa, Italy. ‡‡ Professor and Chairman, §§ Staff Anesthesiologist, Department of Anesthesiology of the Academisch Ziekenhuis Dijkzigt, Rotterdam, the Netherlands. ∥∥ Professor, ## Staff Anesthesiologist, Department of Anesthesiology and General Intensive Care, University of Vienna, Vienna, Austria. *** Members of the Xenon Study Group are listed in the 1.
Article Information
Education
Education   |   January 2003
Multicenter Randomized Comparison of the Efficacy and Safety of Xenon and Isoflurane in Patients Undergoing Elective Surgery
Anesthesiology 1 2003, Vol.98, 6-13. doi:0000542-200301000-00005
Anesthesiology 1 2003, Vol.98, 6-13. doi:0000542-200301000-00005
ALL known general anesthetics, even the modern ones, are described as having side effects. One of the most important side effects of presently used anesthetics is negative inotropicity, which often causes problems in patients with compromised cardiovascular systems. Xenon, currently being investigated as an anesthetic, is interesting as it appears to lack these effects.
In 1951, the first results of the use of xenon as an anesthetic agent in humans undergoing surgery were published. 1 Xenon is described as having many of the characteristics of an ideal anesthetic agent. It is nonexplosive, nonflammable, has low toxicity, and is devoid of known teratogenic effects. Moreover, induction and recovery are rapid because of its blood/gas partition coefficient of 0.115, the lowest of all known anesthetics. 2 While the minimum alveolar concentration (MAC) of xenon, which is considered as a measurement of anesthetic potency and reflects 50% of a population that does not respond to a defined pain stimulus, was first determined as 71%, 3 today it is assumed that the MAC of xenon is 63%. 4 Over the years, several articles were published that covered, in animal studies as well as in a limited number of patients, various aspects of the properties of xenon in anesthesiology. From these studies, xenon, as an anesthetic agent, is believed to have a good safety profile since systemic hemodynamics, the cardiovascular system, and local organ perfusion, especially in the heart, seem not to be affected. 5–8 
Another factor that is gaining increasing public interest is environmental protection. It is known that conventional anesthetic agents such as halogenated alkanes or alkyl ethers, as well as nitrous oxide (N2O), are involved in the destruction of the ozone layer and contribute to the greenhouse effect. This is not the case for xenon, which has no negative environmental effects. Xenon belongs to the group of noble gases (atomic number 54). It is found in very small concentrations in the air (0.0000087%) and is manufactured by a fractional distillation process of liquid air during the process of pure oxygen production. Therefore, xenon anesthesia will not encumber the environment.
Since xenon is expensive and closed circuit anesthesia machines especially modified for xenon application allowing anesthesia with low amounts of gas have not been developed, xenon has only been used in small studies with different experimental settings, including a total of a little more than 100 patients over the last 10 yr. 5,8–10 Therefore, systematic preclinical investigations in a larger group of patients focusing on efficacy and safety of xenon anesthesia are lacking. The aim of this clinical multicenter trial was to test the hypothesis of whether recovery after xenon anesthesia is faster as compared with an accepted, standardized anesthetic regimen and that it is effective and safe. For this study in patients undergoing elective surgery, a new closed circuit anesthetic machine recently modified for xenon application was used.
Methods
The study was designed as a multicenter, randomized, single-blind trial with parallel groups. Six European centers participated. The protocol was approved by all six institutional ethics committees or review boards, and all regulatory requirements of the involved countries were met. All patients gave informed, written consent before participation in this clinical trial.
Patients
A total of 224 patients (target number: 36 patients per center) were enrolled in the trial. Each center recruited their patients in accordance with a pregenerated randomization list (Rancode 3.6 Professional; IDV, Gauting, Germany). The randomization technique used a block design and was generated on a center-by-center basis. The centers received numerated closed envelopes assigning the patients to one of two groups. These envelopes were opened before induction of anesthesia. To accustom the centers to the protocol and the techniques, two to four pilot patients could be treated in each center before starting the trial. The data of these pilot patients were not part of the analysis.
To participate in the study, all of the following inclusion criteria had to be met: age 18 yr or older, American Society of Anesthesiologists (ASA) classification I–III, elective surgery, and planned duration of inhalational anesthesia 2 h or less. During the screening period, the following exclusion criteria were checked: age less than 18 yr, ASA classification IV and V, emergency operation, female patients of childbearing potential without adequate contraception, pregnancy, breast-feeding period, increased intracranial pressure, alcohol or drug abuse, arterial oxygen saturation less than 90% with a fraction of inspired oxygen of 0.21, myocardial infarction within 6 months, stroke within 12 months, disturbed liver function (transaminases threefold over the upper limit), disturbed renal function (serum creatinine twice the upper limit), adrenal insufficiency, congestive heart failure, diabetes mellitus (insulin-dependent), and presumed uncooperativeness or legal incapacity.
A patient could have been withdrawn from the study for the following reasons: withdrawal of consent, lack of efficacy or intraoperative complications such as hypothermia (< 35.5°C), uncooperativeness, noncompliance, serious adverse events, or any reason that, in the eyes of the investigator, did not justify continuation of the study (e.g.  , intercurrent disease).
Protocol
The decision to administer premedication with midazolam was left to the discretion of the investigators. Using the standard monitoring system of each center, anesthesia was induced in all patients using propofol (1–2 mg/kg, up to 5 mg/kg) and sufentanil (0.4 μg/kg). Each patient was denitrogenized with 100% oxygen via  face mask until end expiratory measured oxygen concentration was greater than 90%. Tracheal intubation was facilitated by cisatracurium (0.1–0.2 mg/kg). Maintenance of anesthesia was achieved by xenon (60 ± 5% in oxygen) or isoflurane–N2O (end-tidal concentration, 0.5% isoflurane combined with 60 ± 5% N2O in oxygen) using a closed-circuit anesthesia machine (Physioflex; Draeger, Lübeck, Germany). The concentration of xenon was approximately 95% MAC, while the concentration of isoflurane and N2O represented approximately the same equivalent, with the isoflurane being approximately 40% of MAC and the N2O 55% of MAC. Xenon in medical quality was provided by Messer-Griesheim GmbH (Business Unit, Messer Medical, Krefeld, Germany) in steel cylinders. Each pressure gas cylinder contained 1,000 l. Inspiratory xenon concentration was determined using the thermoconductivity measuring device incorporated in the Physioflex anesthesia machine (accuracy:± 3 vol%), which was calibrated automatically when starting the anesthesia machine. In the case that xenon or N2O dropped below 55% during the inhalational period, the Physioflex system was flushed to displace any accumulated nitrogen previously dissolved in blood and tissues and emerging into the closed circuit of the anesthesia system due to the pressure gradient. Flushing was continued until 60% of xenon or N2O was reached. Total amount of anesthesia gas used was determined by the machine's read out. If maintenance of anesthesia was considered insufficient, as indicated by an increase in systolic blood pressure or heart rate by more than 20% from baseline, sufentanil (10 μg) was intravenously injected; this dose could be repeated every 3 min. In the case that the surgical procedure required muscle relaxation, additional boluses of cisatracurium were administered. At the end of the surgical procedure, patients received neostigmine to reverse muscle relaxation if train-of-four twitch revealed a train-of-four less than 0.7. Antihypertensive, anticholinergic, or inotropic agents could be given during surgery if the heart rate or blood pressure indicated their use despite adequate anesthesia. The study protocol allowed each center to use its standard treatment of blood fluid loss, its standard fluid replacement strategy, and its standard criteria for when to use an arterial line or a central line.
Anesthesia gas was discontinued when all surgical interventions, including the bandaging of the surgical fields, were completed. From this time point, adequate spontaneous ventilation, with an end-expiratory carbon dioxide partial pressure ranging between 40 and 50 mmHg, had to be ensured. After this had been achieved, extubation was performed after the patient opened his or her eyes on command. If opioids were indicated, postoperative pain management was achieved with morphine (0.05–0.1 mg/kg) only.
Assessment of Efficacy
The time between discontinuing the inhalational anesthetic gas and opening the eyes, as well as the time between stopping the gas and extubation, was documented. As the primary endpoint, a recovery index (RI) was used, defined as the ratio between the Aldrete score 11 5 min after extubation plus 1 and the weighted sum of the extubation time (factor 2) plus the time to open eyes on command (factor 1) after the end of anesthesia (= stopping the inhalational anesthetic gas, end of all surgical interventions):MATH
The validity of this index was tested by means of literature data. 12,13 Model calculations revealed stable between-group differences of about 0.17 for both of these independent studies, and both studies favored desflurane versus  isoflurane. As a consequence, a difference exceeding this value can be classified as clinically relevant.
Secondary efficacy endpoints were the time to opening the eyes, time to extubation, and Aldrete score 5, 10, 15, 30, 45, 60, 90, 120, and 240 min after extubation, all assessed by a blinded rater. Furthermore, the cumulative quantity of the administered anesthetic agents as well as hemodynamic (mean arterial pressure, heart rate) and respiratory parameters (arterial oxygen saturation measured by pulse oximetry) were chosen as secondary endpoints.
Assessment of Safety
As safety parameters, preoperative and postoperative 12-lead electrocardiogram, optional electroencephalogram, bispectral index as an additional measure of depth of anesthesia, body temperature, intraoperative need of anticholinergics, antihypertensives, and inotropic substances, and need for postoperative analgesics were recorded. In addition, 5, 10, 15, 30, 45, 60, 90, 120, 240 min, and 24 h after extubation, postoperative pain was measured using the visual analog scale, and vital signs and arterial oxygen saturation were assessed. The time between extubation and readiness for discharge from the recovery room was documented. Laboratory parameters (hematology, clinical chemistry) were measured during the screening phase and 24 h after surgery. Patients were questioned 24 h after extubation to determine whether they recalled anything intraoperatively. Furthermore, all side effects and adverse events were documented. All adverse events or symptoms had to be rated according to severity (mild, moderate, possible, probable, definite). Hypertension was defined as an increase in systolic blood pressure greater than 20% despite three boluses of 10 μg sufentanil concomitant with a stable heart rate and no other signs of low depth of anesthesia, such as sweating. Hypotension and bradycardia were defined as a change of greater than 20% from baseline. Mild hyperthermia was defined as temperatures between 37.5 and 38.5°C. Severe postoperative pain was assumed if patients complained about severe pain or demanded pain medication. Postoperative hypoventilation was indicated by a respiratory rate less than 8 breaths/min. Mental disorientation was noted if a patient was not conscious about his person or whereabouts 30 min after awakening from anesthesia.
Determination of the Sample Size
When taking the clinically relevant advantages between volatile anesthetics into account, it was obvious from the model calculations during the validation in accordance with published material 12,13 that the RI must reach at least a difference of 0.17 for the between-group situation. As a consequence, it was the intention to either accept or reject the Hohypothesis of no difference for this critical value. The sample size calculation revealed a minimum number of 90 patients with an α of 0.001 and a power of 90%.
As the variation remained unclear, it was decided to recruit 216 patients (n = 36 per center). This considerably higher number of patients would give more reliability to safety issues.
Statistics
Data are given as mean ± SD of the intention-to-treat population. The analysis of data ensued after verification of the database by means of the double-entry technique. Thereafter, the data were described descriptively taking the respective scale level into account. Before pooling the data, baseline homogeneity tests were performed. The statistical test methods were performed in accordance with the protocol, i.e.  , confirmatory for the primary target criterion (RI) and exploratively for all secondary criteria to avoid α adjustments. The primary target criterion was tested with the log-rank test in the form of the Peto generalized Wilcoxon test. Secondary criteria were tested with the Wilcoxon–Mann–Whitney U test for between-group situations. For nominal scale and ordinal scale data, the chi-square and Mantel-Haenszel tests were applied, respectively. Statistical significance was indicated by P  < 0.05. A significance level of at least 1% was envisaged for the primary target criterion to either keep or reject the H0hypothesis for no difference.
Results
A total of 218 patients completed the study according to the protocol (36 patients in five centers and 38 in one). In addition to these patients, four patients were withdrawn because of intercurrent events during surgery, and two patients who were wrongly allocated (table 1) were included in the intent-to-treat population (n = 224). The two wrongly allocated patients remained in the correct randomized groups for all parameters except adverse events. The adverse events were reported for the treatment that the patient actually received.
Table 1. Patients Withdrawn during Course of Study; Incorrectly Allocated Patients
Image not available
Table 1. Patients Withdrawn during Course of Study; Incorrectly Allocated Patients
×
The two study groups did not differ with respect to age, weight, height, gender, ASA classification, use of midazolam premedication, and duration of surgery or anesthesia (table 2). In the xenon group (n = 112), the total amount of xenon used per patient was 24.6 ± 10.2 l, and in the isoflurane–N2O group (n = 112), the total amount of isoflurane used per patient was 7.5 ± 3.7 ml. The dose of sufentanil, which was given during anesthesia, did not differ between the two regimens. Postoperatively, patients in the xenon group showed a tendency to need less morphine than patients in the isoflurane–N2O group, without reaching the significance level (16.2 ± 16.8 vs.  21.8 ± 27.6 mg;P  = 0.079).
Table 2. Demographic Data of Intention-to-treat Population
Image not available
Table 2. Demographic Data of Intention-to-treat Population
×
Recovery from Anesthesia
Recovery from anesthesia was faster in the xenon group than in the isoflurane–N2O group (fig. 1). Recovery from xenon anesthesia was independent of the duration of the anesthesia. The RI, used as the primary endpoint, was (0.73 ± 0.38 vs.  0.43 ± 0.28 min−1;P  < 0.0001), indicating a distinctively faster recovery in the xenon group than in the isoflurane–N2O group. According to model calculations before the start of the study, a difference of at least 0.17 was classified as a clinically relevant difference. The time between stopping the inhalational anesthetic gas and opening of the eyes was shorter in the xenon group when compared with the isoflurane–N2O group (4.7 ± 2.3 vs.  8.3 ± 5.4 min;P  < 0.0001), and the time between stopping the inhalational anesthetic gas and extubation was shorter in the xenon group in comparison to the isoflurane–N2O group (5.7 ± 2.8 vs.  9.9 ± 6.0 min;P  < 0.0001). The Aldrete score was always higher in the xenon group than in the isoflurane–N2O group (P  < 0.0001; 5 min after extubation). The time between extubation and readiness for discharge from the recovery room did not differ between groups (144 ± 107 vs.  174 ± 178 min;P  = 0.24).
Fig. 1. Recovery after end of anesthesia given as mean ± SEM. All parameters, which were used to calculate the recovery index, as well as the recovery index itself indicated a faster recovery in the xenon group than in the isoflurane–nitrous oxide group (Wilcoxon–Mann–Whitney U test and log-rank test, *P  < 0.01).
Fig. 1. Recovery after end of anesthesia given as mean ± SEM. All parameters, which were used to calculate the recovery index, as well as the recovery index itself indicated a faster recovery in the xenon group than in the isoflurane–nitrous oxide group (Wilcoxon–Mann–Whitney U test and log-rank test, *P 
	< 0.01).
Fig. 1. Recovery after end of anesthesia given as mean ± SEM. All parameters, which were used to calculate the recovery index, as well as the recovery index itself indicated a faster recovery in the xenon group than in the isoflurane–nitrous oxide group (Wilcoxon–Mann–Whitney U test and log-rank test, *P  < 0.01).
×
Hemodynamic and Respiratory Parameters
Endpoint analyses at the end of the induction period as well as at the end of anesthesia revealed a higher mean arterial pressure for the xenon group in comparison with the isoflurane–N2O group (P  < 0.0001 and P  = 0.001, respectively;fig. 2A). The decrease in heart rate from baseline was more pronounced in the xenon group than in the isoflurane–N2O group (P  = 0.026 and 0.034, respectively;fig. 2B). Intraoperatively, arterial oxygen saturation did not differ between groups (α= 0.05).
Fig. 2. Hemodynamic response to the two anesthetic procedures given as mean ± SEM. (A  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a higher mean arterial pressure for the xenon group in comparison with the isoflurane group (Wilcoxon–Mann–Whitney U test, change from baseline, P  < 0.0001 and P  = 0.001, respectively). (B  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a lower heart rate for the xenon group in comparison with the isoflurane–N2O group (P  = 0.0262 and P  = 0.0341, respectively).
Fig. 2. Hemodynamic response to the two anesthetic procedures given as mean ± SEM. (A 
	) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a higher mean arterial pressure for the xenon group in comparison with the isoflurane group (Wilcoxon–Mann–Whitney U test, change from baseline, P 
	< 0.0001 and P 
	= 0.001, respectively). (B 
	) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a lower heart rate for the xenon group in comparison with the isoflurane–N2O group (P 
	= 0.0262 and P 
	= 0.0341, respectively).
Fig. 2. Hemodynamic response to the two anesthetic procedures given as mean ± SEM. (A  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a higher mean arterial pressure for the xenon group in comparison with the isoflurane group (Wilcoxon–Mann–Whitney U test, change from baseline, P  < 0.0001 and P  = 0.001, respectively). (B  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a lower heart rate for the xenon group in comparison with the isoflurane–N2O group (P  = 0.0262 and P  = 0.0341, respectively).
×
Safety Parameters
Preoperative and postoperative 12-lead electrocardiogram did not reveal relevant changes in any patient of both groups. Postoperatively, none of the patients reported intraoperative awareness. Body temperature decreased slightly in both groups when compared with baseline (xenon group, 36.5 ± 0.3°C vs.  36.3 ± 0.7°C; isoflurane–N2O group, 36.4 ± 0.3°C vs.  36.2 ± 0.6°C) but did not differ between groups (P  = 0.21). No patients in the isoflurane–N2O group, but one patient in the xenon group exhibited four periods of mild hyperthermia, defined as temperatures between 37.5 and 38.5°C. Periods of increased mean systemic arterial blood pressure of more than 20% from baseline despite adequate depth of anesthesia occurred 17 times in the xenon group and 9 times in the isoflurane–N2O group (P  = 0.1431). The intraoperative need for anticholinergics did not differ between groups (xenon, n = 18; isoflurane–N2O, n = 18). The need for inotropic substances was less in the xenon group (xenon, n = 8; isoflurane–N2O, n = 20;P  = 0.0249), whereas the need for antihypertensives was higher in the xenon group (xenon, n = 24; isoflurane–N2O, n = 9;P  = 0.0076). Postoperatively, heart rate, mean arterial pressure, and arterial oxygen saturation remained in the safe range in both groups. Postoperative pain visual analog scale scores did not differ between groups. In general, hematology and clinical chemistry laboratory data of the screening phase and 24 h after surgery did not show any relevant changes that might be caused by the two anesthetic regimens.
The side effect and adverse event profiles did not differ between the two groups (table 3).
Table 3. Number of Occurrence of Adverse Events in Intention–to-treat Population
Image not available
Table 3. Number of Occurrence of Adverse Events in Intention–to-treat Population
×
Discussion
This is the first multicenter trial on the use of xenon as an anesthetic gas, which demonstrates in a large group of ASA I–III patients undergoing elective surgery that anesthesia with xenon can be performed as effectively and safely as with the established anesthesia regimen of isoflurane–N2O. At the same time, xenon anesthesia is associated with a distinctly faster recovery from anesthesia when compared with isoflurane–N2O.
To ensure optimal comparison between the two anesthetics and to eliminate methodical problems, we decided to use as the comparator drug another inhalational anesthetic able to be delivered via  the closed-circuit anesthesia machine. Since desflurane, at present, technically cannot be administered via  the anesthetic circuit we used, and sevoflurane was not available in all study centers, isoflurane, the most widespread inhalational anesthetic, was used. We aimed at the equipotent concentration of both anesthetic regimens of 1 MAC, which is defined as the MAC of an anesthetic that does not cause a response to a defined pain stimulus in 50% of patients. While the concentration of combined used gases, isoflurane (0.50 ± 0.053%) and N2O (59.7 ± 10.16%), resulted in 1 MAC, in the xenon group, 55.6 ± 9.8% xenon was used, which is close to 90% of MAC assuming for xenon a MAC of 63%. 4 Although there is still an ongoing discussion on the real MAC of xenon, it can be assumed that in the xenon group, 1 MAC was not completely achieved and, therefore, only nearly equipotent concentrations were compared. Despite the fact that a little less than 1 MAC was used during xenon anesthesia, it reflects the clinical situation when using xenon in a closed-circuit anesthesia machine. Using a closed circuit causes an accumulation of nitrogen to be dissolved in blood and tissue during normal air breathing and emerging into the closed circuit when ventilated with an oxygen–xenon gas mixture. Two options exist to prevent this accumulation of nitrogen: the first is to breathe pure oxygen for about 20 min before onset of anesthesia to remove the dissolved nitrogen from the body, which is not possible during routine procedures, and the second is to periodically flush the closed system with fresh oxygen–xenon to displace any accumulated nitrogen after partial denitrogenization with 100% oxygen via  face mask during the induction of anesthesia; this was done during the current study. To ensure adequate oxygenation and with respect to the total use of xenon, it was decided to flush the system when the xenon concentration decreased to less than 60%. However, it has to be realized that, in patients requiring more than 30–35% oxygen to maintain an acceptable hemoglobin saturation, the accumulation of nitrogen and the needed oxygen concentration will reduce the xenon concentration to significantly less than 1 MAC.
Both anesthetic regimens produced effective anesthesia with a comparable requirement of the opioid sufentanil. Lachmann et al.  5 compared the efficacy of xenon anesthesia in 20 patients with that of N2O in 20 patients and found that the additional need for the opioid fentanyl during xenon anesthesia was only one fifth of that during N2O anesthesia. This might be explained by a different extent of inhibition of the N  -methyl-d-aspartate receptors, the assumed molecular mechanism underlying the anesthetic effect of xenon and N2O, 14 or by a greater inhibitory effect of xenon on spinal dorsal horn neurons than N2O. 15 In our study, xenon did not result in less requirement of an intraoperative opioid and demonstrated a tendency to a reduced postoperative need of opioids when compared with a combined use of isoflurane and N2O. The reason for this intraoperative finding might be that the regimen of combined use of isoflurane and N2O has two action sites, the enhanced activity of inhibitory γ-aminobutyric acid type A receptors by isoflurane and the inhibition of the N  -methyl-d-aspartate receptors by N2O. However, it remains uncertain whether one or both of the mechanisms described above was responsible for the slightly reduced need of morphine opioids postoperatively in patients in the xenon group.
Xenon causes less cardiovascular response in respect to mean arterial blood pressure than does isoflurane–N2O in the group of healthy patients. In contrast to most anesthetics, xenon does not depress myocardial contractility. Stowe et al.  16 recently published data demonstrating, in isolated guinea pig hearts, that unlike other inhalational anesthetics, xenon does not significantly alter heart rate, atrioventricular conduction time, left ventricular pressure, coronary flow, oxygen extraction, oxygen consumption, cardiac efficiency, and flow responses to bradykinin. Moreover, it was shown that xenon does not depress L-type calcium currents in human atrial myocytes. 17 Even in dogs with experimentally induced cardiomyopathies, xenon does not reduce myocardial contractility. 18 In addition, xenon does not impair the reaction of cardiac muscle bundles to positive inotropic stimulation such as isoproterenol or calcium. 19 The apparent lack of negative effects on the cardiovascular system may make xenon particularly useful for patients with restricted myocardial compliance and in cases where current agents that cause vasodilation and reduce contractility can be potentially very dangerous. Periods of increases in mean systemic arterial blood pressure of more than 20% from baseline despite adequate anesthesia occurred 16 times in the xenon group and 9 times in the isoflurane–N2O group. Although the number of slight blood pressure increases did not significantly differ between the groups, it could be speculated that the slightly higher incidence of mild hypertension might reflect the lack of myocardial depression in the xenon group.
Furthermore, endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a lower heart rate for the xenon group in comparison with the isoflurane–N2O group. The lower heart rate during xenon anesthesia might be a result of the higher blood pressure, resulting in a baroreflex-mediated increase in vagal tone.
To describe the recovery after discontinuation of the anesthetic gas, we used an RI, which takes into account the Aldrete score, the time between stopping the anesthetic gas and extubation, and the time between stopping the anesthetic gas and the opening of the eyes. To determine whether the RI would be an appropriate endpoint, published articles were selected that provided the necessary information either directly or indirectly (through extrapolation). 12,13 Model calculations with the results of these publications revealed a difference in the RI of 0.16 12 and of 0.17 13 for the two tested anesthetics. Both publications favored desflurane compared with isoflurane. The fact that both independent studies led to a comparable between-group difference as well as a clinical conclusion confirms that this RI is an instrument for delivering stable and reliable results. Thus, a difference in RI of 0.17 or higher was classified as a clinically relevant advantage.
As the primary endpoint, this RI, with a difference of 0.30, indicated a distinctively faster recovery in the xenon group than in the isoflurane–N2O group. Recovery from xenon anesthesia was independent of the duration of the anesthesia, which is in accordance with data published by Goto et al.  20 This fast recovery from anesthesia is mainly explained by the low blood/gas partition coefficient of 0.115, which makes xenon the least soluble gas that may be used for anesthesia. Since the additional opioid requirement was the same, it can be excluded that the different recovery was caused by the different amounts of opioids.
The incidence of postoperative nausea and vomiting was reported to be high when xenon is used for anesthesia. 21 However, in this multicenter study in a large group of patients, postoperative nausea and vomiting occurred as often in the xenon group as in the isoflurane–N2O group. The incidence of mild hyperthermia was higher in the xenon group than in the isoflurane–N2O group. None of the patients in the isoflurane group exhibited mild hyperthermia. One patient in the xenon group demonstrated all four observed episodes of mild hyperthermia. In this patient, malignant hyperthermia could not be suspected, since end-tidal carbon dioxide remained below 45 mmHg throughout the study period and arrhythmia was not observed intraoperatively or postoperatively. In general, it is very unlikely that xenon causes malignant hyperthermia since animal studies indicated no malignant hyperthermia–inducing property of xenon. 22,23 However, we are not able to explain these slight increases in body temperature in this patient.
During the 211 ± 102 min of xenon anesthesia, a total amount of 24.6 ± 10.2 l xenon per patient was used. Meanwhile, to reduce the high costs of xenon anesthesia and with respect to its limited availability, a recycling system was recently developed, and technical improvements in the closed-circuit system have been performed, allowing a reduction of the amount of xenon required for anesthesia. Moreover, a further reduction of the needed amount of xenon could be expected if the periodic flushing of the closed system with fresh oxygen–xenon to displace any accumulated nitrogen would be avoided by means of complete denitrogenization before the use of xenon.
In conclusion, based on the data of this first multicenter trial on the efficacy and safety of xenon anesthesia in a large group of patients, the use of xenon gas as an inhalational anesthetic appears to be as effective and safe as established anesthesia with isoflurane–N2O and also allows a faster recovery. Since xenon does not seem to influence the myocardial contractility, 16,17,19 future studies should demonstrate that this new anesthetic allows a better outcome in cardiovascular-compromised patients.
The authors thank Dr. Carsten Pilger, Technician (Messer-Griesheim GmbH, business unit Messer Medical, Krefeld, BRD), Bart Westerkamp, Technician (Physio BV–Draeger, Lübeck, BRD), and Dr. Manfred Raba, Statistician (Pharma Consulting, Oberhaching, BRD) for their support.
Appendix
The Xenon Study Group: Rolf Rossaint, M.D., Professor and Chairman; Matthias Reyle-Hahn, M.D., Staff Anesthesiologist; Klaus Hecker, M.D., Staff Anesthesiologist; Jan Baumert, M.D., Staff Anesthesiologist; and Dinah Eletr, M.D., Staff Anesthesiologist, Department of Anesthesiology, University Hospital of the RWTH Aachen, Germany. Jochen Schulte am Esch, M.D., Professor and Chairman; Jens Scholz, M.D., Professor; Peter H. Tonner, M.D., Professor; Claudia Stapelfeldt, M.D., Staff Anesthesiologist; Christoph Hahn, M.D., Staff Anesthesiologist; Jan Stork, M.D., Staff Anesthesiologist; and Andre Rudkoffsky, Student, Department of Anesthesiology, University Hospital Eppendorf, Hamburg, Germany. Phillippe Scherpereel, M.D., Professor and Chairman; Benoit Vallet, M.D., Professor; Jerome Leclerc, M.D., Staff Anesthesiologist; and R. Nieuviarls, M.D., Staff Anesthesiologist, Department of Anesthesiology and Intensive Care, University Hospital of Lille, Lille, France. Francesco Giunta, M.D., Professor and Chairman; Monica Del Turco, M.D., Staff Anesthesiologist; Forfori Francesco, M.D., Staff Anesthesiologist; and Andrea Ferrari, M.D., Staff Anesthesiologist, Department of Anesthesiology and Intensive Care, University of S. Chiara Hospital, Pisa, Italy. Wilhelm Erdmann, M.D., Professor and Chairman; and Rob Tenbrinck, M.D., Ph.D., Staff Anesthesiologist, Department of Anesthesiology, Academisch Ziekenhuis Dijkzigt, Rotterdam, the Netherlands. Alfons F. Hammerle, M.D., Professor; and Peter Nagele, M.D., Staff Anesthesiologist, Department of Anesthesiology and General Intensive Care, University of Vienna, Vienna, Austria.
References
Cullen SC, Gross EG: The anesthetic properties of xenon in animals and human beings with additional observations on krypton. Science 1951; 113: 580–2Cullen, SC Gross, EG
The blood/gas partition coefficient of xenon may be lower than generally accepted. Br J Anaesth 1998; 80: 255–6NA,
Cullen SC, Eger EI, Cullen BF, Gregory P: Observations on the anesthetic effect of the combination of xenon and halothane. A nesthesiology 1969; 31: 305–9Cullen, SC Eger, EI Cullen, BF Gregory, P
Nakata Y, Goto T, Ishiguro Y, Terui K, Kawakami H, Santo M, Niimi Y, Morita S: Minimum alveolar concentration (MAC) of xenon with sevoflurane in humans. A nesthesiology 2001; 94: 611–4Nakata, Y Goto, T Ishiguro, Y Terui, K Kawakami, H Santo, M Niimi, Y Morita, S
Lachmann B, Armbruster S, Schairer W, Landstra M, Trouborst A, van Daal GJ, Kusuma A, Erdmann W: Safety and efficacy of xenon in routine use as an inhalational anesthetic. Lancet 1990; 335: 1413–5Lachmann, B Armbruster, S Schairer, W Landstra, M Trouborst, A van Daal, GJ Kusuma, A Erdmann, W
Boosma F, Ruprecht J, Man in't Veld AJ, de Jong FH, Dzoljic M, Lachmann B: Hemodynamic and neurohumoral effects of xenon anesthesia. Anesthesia 1990; 11: 40–9Boosma, F Ruprecht, J Man in't Veld, AJ de Jong, FH Dzoljic, M Lachmann, B
Nakata Y, Goto T, Morita S: Comparison of inhalation inductions with xenon and sevoflurane. Acta Anaesthesiol Scand 1997; 41: 1157–61Nakata, Y Goto, T Morita, S
Luttrop HH, Romner B, Perhag L, Eskilsson J, Fredriksen S, Werner O: Left ventricular performance and cerebral hemodynamics during xenon anesthesia: A transesophageal echocardiography and transcranial Doppler sonography study. Anesthesia 1993; 48: 1045–9Luttrop, HH Romner, B Perhag, L Eskilsson, J Fredriksen, S Werner, O
Burov NE, Makeev GN, Potapov YN, Kornienko L: Xenon anesthesia: Clinical manifestation, various techniques, Xenon Anesthesia Today. Proceedings of an Expert Meeting on Xenon Anesthesia. Pisa, Pacini Editore, 1997, pp 55–6
Ferrari A, Erdmann W, Del Tacca M, Formichi B, Volta CA, Ferrari E Bissolotti, Giunta F: Xenon anesthesia: Clinical results and recycling of gas. Appl Cardiopulm Pathophys 1998; 7: 153–5Ferrari, A Erdmann, W Del Tacca, M Formichi, B Volta, CA Ferrari, E Bissolotti, Giunta, F
Aldrete JA: The post-anesthesia recovery score revisited. J Clin Anesth 1995; 7: 89–91Aldrete, JA
Loscar M, Al'loff T, Ott E, Conzen P, Peter K: Aufwachverhalten und kognitive Funktion nach Desfluran und Isofluran. Anaesthesist 1996; 45: 140–5Loscar, M Al'loff, T Ott, E Conzen, P Peter, K
Dupont J, Tavernier B, Ghosez Y, Durneck L, Thevenot A, Moktadir-Chalons N, Ruyffelaerc-Moises I, Declerck N, Scherpereel P: Recovery after anesthesia for pulmonary surgery: Desflurane, sevoflurane and isoflurane. Br J Anaesth 1999; 82: 355–9Dupont, J Tavernier, B Ghosez, Y Durneck, L Thevenot, A Moktadir-Chalons, N Ruyffelaerc-Moises, I Declerck, N Scherpereel, P
Franks NP, Dickinson R, de Sousa SL, Hall AC, Lieb WR: How does xenon produce anesthesia? (letter). Nature 1998; 396: 324Franks, NP Dickinson, R de Sousa, SL Hall, AC Lieb, WR
Miyazaki Y, Adachi T, Utsumi J, Shichino T, Segawa H: Xenon has a greater inhibitory effect on spinal dorsal horn neurons than nitrous oxide in spinal cord transected cats. Anesth Analg 1999; 88: 893–7Miyazaki, Y Adachi, T Utsumi, J Shichino, T Segawa, H
Stowe DF, Rehmert GC, Wai-Meng Kwok, Weigt HU, Georgieff M, Bosnjak ZJ: Xenon does not alter cardiac function or major cation currents in isolated guinea pig hearts or myocytes. A nesthesiology 2000; 92: 516–22Stowe, DF Rehmert, GC Wai-Meng, Kwok Weigt, HU Georgieff, M Bosnjak, ZJ
Hüneke R, Jüngling E, Skasa M, Rossaint R, Lückhoff A: Effects of the anesthetic gases xenon, halothane, and isoflurane of calcium and potassium currents in human atrial cardiomyocytes. A nesthesiology 2001; 95: 999–1006Hüneke, R Jüngling, E Skasa, M Rossaint, R Lückhoff, A
Hettrick DA, Pagel PS, Kersten JR, Tessmer JP, Bosnjak ZJ, Georgieff M, Warltier DC: Cardiovascular effects of xenon in isoflurane-anesthetized dogs with dilated cardiomyopathy. A nesthesiology 1998; 89: 1166–73Hettrick, DA Pagel, PS Kersten, JR Tessmer, JP Bosnjak, ZJ Georgieff, M Warltier, DC
Schroth SC, Schotten U, Alkanoglu O, Reyle-Hahn MS, Hanrath P, Rossaint R: Xenon does not impair the reaction of cardiac muscles bundles to positive inotropicic stimulation. A nesthesiology 2002; 96: 422–7Schroth, SC Schotten, U Alkanoglu, O Reyle-Hahn, MS Hanrath, P Rossaint, R
Goto T, Saito H, Nakata Y, Uezono S, Ichinose F, Morita S: Emergence times from xenon anesthesia are independent of the duration of anesthesia. Br J Anaesth 1997; 79: 595–9Goto, T Saito, H Nakata, Y Uezono, S Ichinose, F Morita, S
Petersen-Felix S, Luginbuhl M, Schnider TW, Curatolo M, Arendt-Nielsen L, Zbinden AM: Comparison of the analgesic potency of xenon and nitrous oxide in humans evaluated by experimental pain. Br J Anaesth 1998; 81: 742–7Petersen-Felix, S Luginbuhl, M Schnider, TW Curatolo, M Arendt-Nielsen, L Zbinden, AM
Baur CP, Klingler W, Jurkat-Rott K, Froeba G, Schoch E, Marx T, Georgieff M, Lehmann-Horn F: Xenon does not induce contracture in human malignant hyperthermia muscle. Br J Anaesth 2000; 85: 712–6Baur, CP Klingler, W Jurkat-Rott, K Froeba, G Schoch, E Marx, T Georgieff, M Lehmann-Horn, F
Froeba G, Marx T, Pazhur J, Baur C, Baeder S, Calzia E, Eichinger HM, Radermacher P, Georgieff M: Xenon does not trigger malignant hyperthermia in susceptible swine. A nesthesiology 1999; 91: 1047–52Froeba, G Marx, T Pazhur, J Baur, C Baeder, S Calzia, E Eichinger, HM Radermacher, P Georgieff, M
Fig. 1. Recovery after end of anesthesia given as mean ± SEM. All parameters, which were used to calculate the recovery index, as well as the recovery index itself indicated a faster recovery in the xenon group than in the isoflurane–nitrous oxide group (Wilcoxon–Mann–Whitney U test and log-rank test, *P  < 0.01).
Fig. 1. Recovery after end of anesthesia given as mean ± SEM. All parameters, which were used to calculate the recovery index, as well as the recovery index itself indicated a faster recovery in the xenon group than in the isoflurane–nitrous oxide group (Wilcoxon–Mann–Whitney U test and log-rank test, *P 
	< 0.01).
Fig. 1. Recovery after end of anesthesia given as mean ± SEM. All parameters, which were used to calculate the recovery index, as well as the recovery index itself indicated a faster recovery in the xenon group than in the isoflurane–nitrous oxide group (Wilcoxon–Mann–Whitney U test and log-rank test, *P  < 0.01).
×
Fig. 2. Hemodynamic response to the two anesthetic procedures given as mean ± SEM. (A  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a higher mean arterial pressure for the xenon group in comparison with the isoflurane group (Wilcoxon–Mann–Whitney U test, change from baseline, P  < 0.0001 and P  = 0.001, respectively). (B  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a lower heart rate for the xenon group in comparison with the isoflurane–N2O group (P  = 0.0262 and P  = 0.0341, respectively).
Fig. 2. Hemodynamic response to the two anesthetic procedures given as mean ± SEM. (A 
	) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a higher mean arterial pressure for the xenon group in comparison with the isoflurane group (Wilcoxon–Mann–Whitney U test, change from baseline, P 
	< 0.0001 and P 
	= 0.001, respectively). (B 
	) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a lower heart rate for the xenon group in comparison with the isoflurane–N2O group (P 
	= 0.0262 and P 
	= 0.0341, respectively).
Fig. 2. Hemodynamic response to the two anesthetic procedures given as mean ± SEM. (A  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a higher mean arterial pressure for the xenon group in comparison with the isoflurane group (Wilcoxon–Mann–Whitney U test, change from baseline, P  < 0.0001 and P  = 0.001, respectively). (B  ) Endpoint analysis at the end of the induction period as well as at the end of anesthesia revealed a lower heart rate for the xenon group in comparison with the isoflurane–N2O group (P  = 0.0262 and P  = 0.0341, respectively).
×
Table 1. Patients Withdrawn during Course of Study; Incorrectly Allocated Patients
Image not available
Table 1. Patients Withdrawn during Course of Study; Incorrectly Allocated Patients
×
Table 2. Demographic Data of Intention-to-treat Population
Image not available
Table 2. Demographic Data of Intention-to-treat Population
×
Table 3. Number of Occurrence of Adverse Events in Intention–to-treat Population
Image not available
Table 3. Number of Occurrence of Adverse Events in Intention–to-treat Population
×